Team Status Report for 4/6/24

Risks:

• 2GB Jetson Nano lost support in 2023, and so some of the software is fairly out of date (For example, Python 3.6 was the latest version available to be installed through apt, however Python 3.10 is the minimum version for YOLOv8)

Changes:

• Switched back from classification model to object detection YOLO model because the classification model was buggy and so we switched to a smaller, custom dataset
• Switched from Xilinx Kria KV260 to Nvidia Jetson Nano as server hardware because we weren’t able to accelerate the YOLO model like originally planned. Instead we will be relying on the CUDA cores (which the YOLO model already has acceleration for).

Progress:

• Finished with main App UI
  • Recipe confirmation pop up
  • Recipe card
  • Upload Picture widget
  • Custom inventory input
  • Custom recipe input
  • LLM Integration, using filters
• Created working Server for Demo
  • MySQL Database installed
  • Backend API program runs on local machine
    • Connects clients to the MySQL DB

Verification & Validation:

• Validate CV code makes proper API calls
  • Test that objects being entered send an “Put Item” call
  • Test that objects being removed sends a “Delete Item” call
• Validate WebApp makes proper API calls
  • Test that requesting a recipes sends a “Get All Items” call
  • Test that inventory pages send “Get Items {Location}” call
• Validate API calls update the database
  • Test relevant calls perform relevant actions to database